Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: jordan86 on February 16, 2021, 06:31:54 pm
-
Curious how I might calculate the output resistance of different tubes in my SE amp, knowing the tube type and my voltages? Is there a simple formula for that? I apologize if this has covered. I did my due diligence to search. I thought at one point I fond a thread with a simplified formula like plate voltage x plate voltage / plate dissipation. Can't find that now.
I ask because my amp can run KT66, EL34, 6L6, KT88...etc. I know those probably all have different resistances but data sheets are often not listed in close values. For instance, the Gold Lion KT66 data sheet only has a 250V list for a single valve in class a pentode. My amp runs around 350-400 depending on the tube type and bias. https://www.thetubestore.com/lib/thetubestore/genalex-kt66.pdf
My OT has a 2.5K primary to 4,8, and 16 ohm secondary taps. It does have a 4K and 8K primary option as well but they were cut to the nub so trying to reverse engineer the math at 2.5K to find the closest matches based on the winding ratios.
Edit: I used the word impedance but changed it now. I think resistance would be more accurate in discussing the tubes output.
-
R = V/I and P = V x I
So, turn the 'R' into 'impedance' to get:
Zout = V/I, and rearrange for known factors, V and Pmax, hence:
Zout = V/(P/V), where V = Plate voltage at idle, and P = Maximum plate dissipation at idle
This will find the optimum* 'centre-bias' load resistance for a given tube at a given bias point.
E.g. for a 6V6 operating at 340V
340/(12/340) = 9,633R (or 9k6)
(*I'm assuming you want to get the most mileage out of any particular tube)
If the screen voltage is so high as to end up putting the load line below the 'knee' of the Vg = 0 curve, then either:
1) lower the screen voltage until the load line passes through the knee. This will help limit excessive screen grid dissipation under big signal conditions. (Many Fender SE amps had Vg2 sitting below Vp - especially as the voltages got higher - which helps to achieve this), or
2) decrease the load resistance slightly e.g. to 8k (which rotates the load line so that it crosses at or above the knee of Vg = 0), or
3) Cool off the bias so that the plate and screen aren't working at maximum Pdiss , or
4) A combination of 1), 2) and 3) above
-
Thanks Tubeswell! When calculating dissipation, I assume I measure cathode to plate? So 370 instead of 400v (cathode voltage being 30v to ground). Do I use that 370v value as well for my plate voltage in the equation, or the 400v measurement from plate to ground?
-
Thanks Tubeswell! When calculating dissipation, I assume I measure cathode to plate? So 370 instead of 400v (cathode voltage being 30v to ground). Do I use that 370v value as well for my plate voltage in the equation, or the 400v measurement from plate to ground?
By all means use 370 - you'll end up with a slightly lower Zout, but no harm done. I take it this is a 6L6-type tube?
-
Will do.
Its a Marshall superlead preamp. I have been using the Gold Lion KT66 lately. Has the right balance of top end and compression to my ear. But also have a few EL34’s and KT88’s that I swap on occasion. Realizing the 2.5K is low for all of them. The tones not bad, just the math bugs me. I know it shouldn’t though.
-
What you want for loading is not really a property of the tube but the specific V and I you work it at. (Which in turn may be constrained by tube properties; can't get 500mA through an EL84, etc.)
So for about the same voltage and power, you can put about any big-enough tube with the same load, just adjusting the bias and drive.
If you are not trying to fill a stadium for only $45, I don't see a problem with 2.5k on all the big tubes.
You want to be down near 280V though. 280V at 112mA is over 31 Watts which is enough (maybe too much) for anything under 6550/KT88.
370V and 30W Pdiss means 80mA and 370/80 is 4.6k, so you are in "5k" territory. Even a 42W tube works out as 3.2k.
-
Yeah my kt88 calculates to be about 3.3k, so even my biggest tube doesn’t match, still too high. My 25 watt tubes are in the neighborhood of 5-5.5k, as I had been guessing. I’ve been just running them in a lower OT tap. As in 2.5k primary to 8 ohm, with my 5k tube going to a 16 ohm speaker. So corrected to match.
I did some recording and A/B though yesterday and kinda like the KT66 out of the 16 ohm tap to a 16 ohm load. So that’s really 32 ohms down to a 16 ohm speaker. I guess Matchless did that too. An intentionally low primary for a low mismatch. Just want to make sure I’m not gonna break anything...which tubeswell prob picked up on in his recommendation on reply #1
-
... KT66 out of the 16 ohm tap to a 16 ohm load. So that’s really 32 ohms down to a 16 ohm speaker. I guess Matchless did that too. An intentionally low primary for a low mismatch. ...
I'm not sure what the above means. What is "really 32Ω"?
The speaker is the only thing with an actual fixed impedance here (even then it's not really "fixed" and changes with frequency). The output transformer takes whatever load impedance is present at the secondary and makes it "look like" something else, in proportion to the number of turns on each winding.
So the tube sees a load the the transformer levers-up from whatever speaker/resistive/reactive load is attached at a secondary.
"Mismatch" doesn't matter much, except power is down and distortion changes, until you reach some kind of edge-case where it may be possible to overheat a tube plate or screen.